Séminaire organisé par Hugues de Thé (Inserm U944, Hôpital St Louis, Paris, France), Pier Paolo Pandolfi (Harvard Medical School, Beth Israel Deaconesss Medical Center, Boston, USA) et Leonard Zon (Harvard Medical School, Boston Children’s Hospital, Boston, USA) du 26 septembre au 1er octobre 2016
Ali Bazarbachi, Olivier Bernard, Elaine Dzierzak, Michael Hemann, Raphaël Itzykson, Valérie Lallemand-Breitenbach, Ross Levine, A. Thomas Look, Hind Medyouf, Simon Mendez-Ferrer, Benjamin Neel, Claus Nerlov, Pier Paolo Pandolfi, Emmanuelle Passegué, Pier Giuseppe Pelicci, Jean Soulier, Hugues de Thé, Eirini Trompouki, Leonard Zon
La finesse du démembrement moléculaire des leucémies et des lymphomes et la relative simplicité des anomalies génétiques qui sous-tend leur développement les a presque transformés en maladies rares, voire orphelines. La physiopathologie des plus simples d’entre elles commence à être bien comprise, ouvrant la porte soit à des thérapeutiques ciblées, soit à des stratégies rationnelles exploitant les faiblesses intrinsèques des cellules transformées. Dans ce contexte, les participants au séminaire «Modèles animaux des proliférations hématologiques» ont pu présenter leurs travaux et échanger dans l’atmosphère si féconde des Treilles. Les invités de la Fondation, d’âge, de spécialité et d’origine géographique très différents ont pu interagir très étroitement et confronter leurs visions et leurs expériences. Les premières présentations ont abordé les problèmes de la différenciation et de l’auto-renouvellement des cellules souches sanguines normales, y compris leur vieillissement (Zon, Dzierzak, Trompouki, Passegué). La leucémogenèse procède souvent d’une dérégulation des voies maîtresses régulant les cellules normales. D’autres intervenants ont présenté leurs travaux sur les mécanismes moléculaires à travers lesquels les oncogènes induisent la transformation cellulaire (Nerlov, Levine, Bernard, Pandolfi). Le rôle des facteurs de transcription, des kinases, des phosphatases ou des effecteurs épigénétiques, a été étudié dans des modèles de souris, mais aussi de poisson-zèbre (Look, Zon). La séquence ultra-profonde permet maintenant d’aborder la dérive génétique et l’hétérogénéité tumorale (Soulier). Les interactions cellulaires entre le microenvironnement stromal et les cellules leucémiques ont été abordées par Look, Medyouf et Mendez-Ferrer. Le rôle modulateur des éléments biochimiques externes (stress oxydant, acides gras) a été développé par Pelicci et Neel. L’utilisation des modèles animaux à des fins pré-cliniques a été abordée par la plupart des intervenants. Ceux-ci ont souligné que si la transposition clinique des résultats issus de ces modèles posait parfois des difficultés, elle restait une source irremplaçable d’information pour la compréhension des mécanismes moléculaires des traitements, y compris conventionnels (Hemann, Itzykson, Bazarbachi). Dans certain cas ces modèles ont déjà permis l’optimisation des thérapeutiques (de Thé, Lallemand, Bazarbachi).
The participants of the meeting mouse models of haematopoietic malignancies included physicians, clinician-scientists and basic scientists of very different stages in their careers, institutions and age (two of them were juniors, most others established scientists). In a very fecund atmosphere, normal haematopoietic stem cell biology was discussed, as well as the alterations in the normal mechanisms of self-renewal that underlies development of some leukaemia. The function of other oncogenes, which are not master regulators of stem cell differentiation or self-renewal, was also discussed, as well as the modelling of their function in mouse or zebrafish. Other participants discussed the role of the stromal micro-environment on stem cell transformation. A significant part of the presentations and discussions dealt with therapeutic issues, using animal models. Clinicians extensively discussed issues of relevance of animal models to patients care and protocol design.
Elaine Dzierzak: A wide diversity of hematopoietic cells exists in the mammalian embryo during the generation of the adult hematopoietic system, including that of potent hematopoietic stem cells (HSC). HSC emerge from the major vasculature of the embryo through a transition of endothelial cells to HSC fate. HSC heterogeneity (based on HSC responsiveness to growth factors such as BMP4 and Hedgehog) is established already in the foetal liver. The differences in growth factor responsiveness of HSC types have implications for drug treatment of leukaemia. The Gata2 transcription factor is required for the generation of HSC from embryonic endothelial cells. In a time-lapse vital imaging approach, we showed that Gata2 expression dynamics is highly oscillatory during endothelial-to-hematopoietic cell-transition. Hematopoietic defects in Gata2 haplo-insufficient embryos correlated with altered Gata2 oscillations, implicating a new level of dynamic regulation that may influence the acquisition of HSC fate in health and leukaemia.
Emmanuelle Passegue: With age, hematopoietic stem cells (HSCs) lose their ability to regenerate the blood system, and promote disease development. Autophagy is associated with health and longevity, and is critical for protecting HSCs from metabolic stress. We now show that loss of autophagy in HSCs causes accumulation of mitochondria and an oxidative phosphorylation (OXPHOS)-activated metabolic state, which drives accelerated myeloid differentiation mainly through epigenetic deregulations, and impairs HSC self-renewal activity and regenerative potential. Strikingly, the majority of HSCs in aged mice share these altered metabolic and functional features. However, ~ 1/3 of aged HSCs exhibit high autophagy levels and maintain a low metabolic state with robust long-term regeneration potential similar to healthy young HSCs. Our results demonstrate that autophagy actively suppresses HSC metabolism by clearing activated mitochondria to maintain quiescence and stemness, and becomes increasingly necessary with age to preserve the regenerative capacity of an old HSC compartment.
Len Zon: Haematopoiesis involves the interaction of stem cells within the niche. We have undertaken a technique called Tomo-seq to study cells in the niche. We have found markers that are specific to macrophages that interact with stem cells, and endothelial cells. The endothelial cells of the venous sinusoids of the niche express E-selectin. We also have shown that PGE2 can stimulate blood stem cells to acquire a memory. This appears to be due to an epigenetic mechanism. Our studies have shown other lipids like EETS that can stimulate stem cells to migrate. Together, our work shows the interactions of specific types of cells in the niche with stem cells to stimulate alterations in cell behaviour.
Eirini Trompouki: Hematopoietic Stem Cells (HSCs) provide the foundation of the hematopoietic system in vertebrates. Being multipotent and capable of self-renewal, HSCs are responsible for constant production of all blood cell types throughout life. However, deregulation in HSCs’ program leads to a panel of hematopoietic malignancies that, in many cases, exhibit reactivation of developmental pathways and can be easily studied in zebrafish. During the meeting we discussed the role of the transcription factor HLX in hematopoietic development. This factor is overexpressed in Acute Myeloid Leukaemia patients, as identified by the Steidl laboratory. Our data suggest that HLX positively regulates HSC formation by fine-tuning metabolic pathways. Additionally, we referred to a family of RNA helicases, implicated in immune diseases, and discussed how they act as a network to modulate HSC development and maintenance.
Klaus Nerlov: Acquired mutations in the CEBPA gene are observed in around 10% of acute myeloid leukaemia. To understand how CEBPA mutations initiate the disease we have knocked both major classes of human CEBPA mutation into the mouse germ line. C-terminal CEBPA mutations (which lack DNA binding) were observed to expand premalignant HSCs, but could not support the formation of leukemic stem cells due to a block in myeloid lineage commitment. In contrast, N-terminal CEBPA mutations (which block expression of the p42 C/EBPalpha isoform) supported the formation of committed myeloid progenitors, but not premalignant HSC expansion. By combining the two mutations, the most common occurrence in de novo AML patients, both processes were supported. Accordingly, leukaemo-genesis was accelerated in the presence of one C- and one N-terminal mutation compared to both homozygous situations. These findings provide a cellular basis for the clinically observed pattern of CEBPA mutations.
Olivier Bernard: Two lines of results have been presented. The first one describes studies of the oncogenic cooperation between TET2 and DNMT3A mutants. Both genes encode proteins regulation DNA methylation and are frequently mutated in human haematological malignancies of both myeloid and lymphoid subtypes. The mouse model recapitulates the human pathologies and shows that the oncogenic cooperation arises from additive deregulation of the DNA methylation control. The second line showed that somatic mutations observed in chronic lymphocytic leukaemia (CLL) patients are present in progenitor cells able to differentiate along both lymphoid and myeloid lineages. These observations might have consequences in the outcome of novel treatments inhibiting BCR signalling that target mature leukemic cells while likely leaving unarmed mutated progenitors.
Jean Soulier presented data on clonal evolution and leukaemogenesis in patients with Fanconi anemia, in their constitutive context of genomic instability and HSCs exhaustion. The work reveals the progressive emergence of cells with a highly recurrent pattern of somatic abnormalities that are mostly related to unbalanced chromosomal translocations and that lead to partial chromosomal arm duplications or losses. Sequencing many translocation breakpoints showed molecular signatures of end joining repair rather than homologous recombination, which suggests the ectopic use of these systems to drive clonal evolution. The most frequent and early somatic lesion in FA are partial duplication of chromosome 1q (1q+, present in more than half the cases of MDS/AML), following by 3q+, inactivation of RUNX1, and deletion of 7q, the target genes and leukemogenic functions of which have been investigated and experimentally modelled. This work shed enable to draw a model for specific leukemia progression in the context of a constitutive DNA repair defect.
Ross Levine: The discovery of JAK2/MPL mutations in the majority of patients with Myeloproliferative Neoplasms (MPN) led to the development of JAK kinase inhibitors. JAK kinase inhibitors, including ruxolitinib, improve patient constitutional symptoms and reduce splenomegaly, but do not significantly reduce mutant allele burden. Chronic exposure to JAK kinase inhibitors results in JAK inhibitor persistence in cell lines, murine models, and patient samples, which we have shown is due to JAK2-transactivation and persistent JAK-STAT signalling. We have used murine genetic studies show that MPN cells remain dependent on JAK2 expression for proliferation and survival. These data suggest that pharmacologic approaches that better inhibit JAK2 activity in MPN cells may demonstrate increased efficacy in vivo. In addition, recent studies have shown that a subset of MPN patients has concurrent mutations in epigenetic regulators, most commonly in the chromatin modifier ASXL1, and that concurrent JAK2/ASXL1 mutations are associated with adverse outcome in MF patients. Our data underscore the need for new therapeutic approaches for MF patients based on mechanistic insight into disease pathogenesis. We presented data on the mechanisms by which JAK2 and ASXL1 mutations cooperate to induce myeloid transformation, to drive disease progression and adverse outcome, and how concurrent mutations in JAK2 and ASXL1 affect the response to JAK kinase inhibitors.
Dr. Pandolfi discussed his recent findings on the PTEN tumor suppressor.
PTEN is a critical tumor suppressor phosphatase, active in its dimer configuration within membrane compartments. However, the mechanisms regulating PTEN dimerization are currently. unknown. Dr. Pandolfi finds that an atypical type of poly-ubiquitination suppresses PTEN dimerization, membrane recruitment and function. They show that the E3 ligases that mediate this modification interact with PTEN and are essential to cooperatively catalyze this non-degradative modification. They further show that these E3 ligases are direct MYC target genes and are critical for its tumorigenic function. Importantly, they found that the pharmacological inhibition of these ligases triggers PTEN reactivation and a potent suppression of MYC-driven tumorigenesis both in vitro and in vivo. Their novel findings therefore unravel the oncogenic role for atypical PTEN poly-ubiquitination, and a therapeutic strategy for the treatment of MYC-driven cancers through PTEN reactivation.
Benjamin Neel: described two recent new developments that revealed unexpected pathways regulated by two members of the protein-tyrosine phosphatase family, PTPN1 (encoding PTP1B) and PTPN11 (encoding SHP2). First, he reviewed work identifying the PTP1B/RNF213/a-ketoglutrate dioxygenase (aKGDD) pathway, which is required for survival of breast cancer cells in severe hypoxia. PTP1B appears to dephosphorylate the large AAA-ATPase/E3 ligase RNF213, which in turn regulates aKGDD globally, possibly viva control of intracellular ascorbate (vitamin C levels). He also discussed a specific requirement for SHP2 to enable proliferation of BCR/ABL1-transformed, but not normal, pre-B cells. SHP2 is required for the activation of both the ERK MAPK and for Src family PTKs in BCR/ABL-transformed cells. In normal pre-B cells, however, SHP2 is only required for ERK activation, and ERK activation is dispensable for their proliferation. Downstream of these events, SHP2 deficiency appears to affect a distinct subset of MYC regulated genes, perhaps by controlling the transcription of MXD3,4, two key negative regulators of MYC-dependent gene expression.
Thomas Look: discussed newly developed zebrafish models of human myeloid malignancies. These fish develop MDS or MPN after the orthologue of the human suppressor genes, TET2, ASXL1 and DNMT3A have been inactivated in the zebrafish genome by genomic editing using CRISPR-cas9. These fish produce hundreds of embryos per week and Dr Look’s team is taking advantage of the conserved blood system and optical clarity of these fish to analyse large collections of drugs approved for human. This work will identify those that are selectively toxic to hematopoietic stem cells harbouring these mutations. This approach is called « repurposing » and drugs identified in this way can be quickly mobilized into clinical medicine. Dr Look’s laboratory has identified several promising drugs using this approach and they are now being tested in murine models to verify activity in mammalian blood cells. This is the last step in pre-clinical analysis before they can be considered for testing in clinical trials of patients harbouring these mutations.
Hugues de Thé presented an historical overview of his APL studies and stressed how therapy by retinoic acid and arsenic trigger not only APL eradication, but a well-characterized cascade of events that culminate with PML/P53 driven senescence. Molecular details were presented and discussed, including analysis of resistant or relapsing patients that fully support the model established in the mouse. Some unorthodox findings on RARA-mediated immortalization of haematopoietic progenitors were also presented.
Valérie Lallemand-Breittenbach presented a set of experiments aimed at understanding the function of the PML tumour suppressor in non-APL cells. She presented cell biology analyses showing how oxidative stress underlies PML polymerization and assembly of nuclear bodies. Conversely, PML is a sensor for oxidative stress in normal mice and controls ROS-induced P53 activation. She also presented data highlighting the actual role of sumoylation in PML assembly, degradation and partner recruitment.
Pier Giuseppe Pelicci is studying the effects of environmental signals on leukaemia and mammary tumour stem cells (SCs). In the first part of the talk (effect of the microenvironment), I will present our recent results on the role of the oncogene myc in cancer SC expansion via increasing symmetric divisions and progenitor reprogramming, and discuss the contribution of components of the immune-surveillance to the elimination of DNA-damaged SCs and the underlying role of p21. In the second part (effect of the macro-environment), I will discuss the effect of Caloric restriction (CR) on leukaemia SCs in a mouse model of Acute Myeloid Leukaemia and the potential value of therapeutic approaches such as the combination of LSD1 inhibition with CR or Insulin Inhibitors to eradicate disease. Finally, I will present some of our results on the link between obesity and cancer risks in the presence of FLT3-ITD mutations.
Hind Medyouf focused on the importance of cellular cross-talk between bone marrow niche cells and their hematopoietic counterpart, in the context of human myelo-dysplasia (MDS). The presentation provided deeper insight into the evolutionary history of MDS development and how these pre-leukemic cells interact and modify their microenvironment to promote disease maintenance/progression. Deciphering these key cellular interactions will hopefully allow us to devise better therapeutic strategies for patients.”
Simon Mendez-Ferrer: Myeloid malignancies are caused by genetic or epigenetic alterations in haematopoietic stem cells (HSCs) and/or myeloid progenitor cells (MPCs). They comprise myelo-proliferative neoplasms (MPNs), myelodysplastic syndromes and acute myeloid leukaemia (AML). MPN patients have a higher risk of developing AML. HSCs/MPCs can transform into leukaemic stem cells (LSCs), which are chemo-resistant and cause tumour relapse. Increasing evidence and our own data support a key role of the microenvironment or “niche” where LSCs reside during leukaemo-genesis and resistance. Our past work identified bone marrow mesenchymal stem cells (BMSCs) with essential HSC niche functions and demonstrated that neuropathy and BMSC apoptosis contribute to MPN progression, suggesting potential ways to treat MPN by targeting the niche. However, we unpublished data presented at this meeting shows striking differences in the role of BMSCs in MPN and AML, arguing for the need of stage-specific interventions on the micro-environment.
Mickael Hemann discussed the basis for the activity of platinum derivatives in cancer therapy. In particular, he provided evidence that some derivative do not kill cancer cells through induction of DNA damage. In particular, oxaliplatin kills cells by inducing ribosome biogenesis stress. These studies highlight the urgent need for basic biology approaches of to reassess the actual bases for conventional chemotherapy efficacy.
Raphael Itzykson investigated the role of the BET family of epigenetic readers in the leukaemia-propagating capacity of AML cell lines and primary cells. We found that chemical inhibition of BET proteins at concentrations that preserve the viability of leukemic cells is nonetheless able to reduce clonogenic output in different oncogenic contexts. We identified a core network of 8 BET-regulated genes associated to this leukaemia-propagating phenotype including known MYC-targets involved in metabolic pathways such as one-carbon mitochondrial pathway. Using long-term culture of primary AML cells, and xenotransplantation validation, we also identified oncogenic subgroups of AML where BET inhibition reduces leukaemia-propagating activity.
Ali Bazarbachi: ATL is a chemotherapy-resistant malignancy caused by HTLV-I infection and associated with dismal prognosis. The viral oncoprotein Tax initiates ATL in transgenic mice. The combination of arsenic trioxide (arsenic) and interferon-α (IFN) triggers proteasome mediated Tax degradation and apoptosis in ATL cells and cures ATL in mice through specific targeting of leukaemia initiating cells (LIC). These preclinical results led to encouraging clinical improvement in ATL patients when treated with arsenic/zidovudine/IFN. Beyond ATL, this represents a unique and novel model of studying treatment-induced LIC eradication following therapy-induced loss of the driving oncogene